Science of the Total Environment 511 (2015) 703–710

Contents lists available at ScienceDirect

Science of the Total Environment journal homepage: www.elsevier.com/locate/scitotenv

Using chemical, microbial and fluorescence techniques to understand contaminant sources and pathways to wetlands in a conservation site J. Rhymes a,⁎, L. Jones b, D.J. Lapworth c, D. White c, N. Fenner a, J.E. McDonald a, T.L. Perkins a a b c

Bangor University, UK Centre for Ecology and Hydrology, UK British Geological Survey, UK

H I G H L I G H T S • • • •

We used chemistry, fluorescence and bacterial counts to study multiple pathways of contamination to a wetland site. Contamination sources are primarily fertilizers, causing exceedance of nutrient thresholds in groundwater within the site. Contamination pathways into the site include groundwater, surface runoff, and streams, with minor input from on-site grazers. The site attenuates nutrient and bacterial concentrations providing an ecosystem service, but with adverse biological effects.

a r t i c l e

i n f o

Article history: Received 5 November 2014 Received in revised form 23 December 2014 Accepted 24 December 2014 Available online 20 January 2015 Editor: D. Barcelo Keywords: Faecal indicator bacteria Dissolved organic matter fluorescence Water chemistry Groundwater Sand dunes Dune slacks Ecosystem service

a b s t r a c t Nutrients and faecal contaminants can enter wetland systems in a number of ways, with both biological and potentially human-health implications. In this study we used a combination of inorganic chemistry, dissolved organic matter (DOM) fluorescence and Escherichia coli and total coliform (TC) count techniques to study the sources and multiple pathways of contamination affecting a designated sand dune site of international conservation importance, surrounded by agricultural land. Analysis of stream samples, groundwater and dune slack wetlands revealed multiple input pathways. These included riverbank seepage, runoff events and percolation of nutrients from adjacent pasture into the groundwater, as well as some on-site sources. The combined techniques showed that off-site nutrient inputs into the sand dune system were primarily from fertilisers, revealed by high nitrate concentrations, and relatively low tryptophan-like fulvic-like ratios b 0.4 Raman units (R.U.). The E. coli and TC counts recorded across the site confirm a relatively minor source of bacterial and nutrient inputs from on-site grazers. Attenuation of the nutrient concentrations in streams, in groundwater and in run-off inputs occurs within the site, restoring healthier groundwater nutrient concentrations showing that contaminant filtration by the sand dunes provides a valuable ecosystem service. However, previous studies show that this input of nutrients has a clear adverse ecological impact. © 2015 Elsevier B.V. All rights reserved.

1. Introduction The global availability and mobility of nitrogen have increased rapidly over the past five decades (Galloway and Cowling, 2002) and the damaging impacts it has on freshwater ecosystems are widely documented (Camargo and Alonso, 2006). Aquatic systems are extremely sensitive to nitrogen and are threatened by atmospheric deposition inputs (Fowler et al., 2005) as well as point sources and diffuse sources which can enter aquatic systems via numerous pathways such as through runoff, streams and groundwater. Tracing sources of aquatic pollution is therefore often problematic (Withers et al., 2009).

⁎ Corresponding author. E-mail address: [email protected] (J. Rhymes).

http://dx.doi.org/10.1016/j.scitotenv.2014.12.085 0048-9697/© 2015 Elsevier B.V. All rights reserved.

Within rural areas river water quality (Hooda et al., 2000) and groundwater quality (Oakes et al., 1981) are primarily impacted by agricultural diffuse pollution (Novotny, 1999). Atmospheric nutrients have been demonstrated to have adverse impacts on the ecology of protected dune habitats (Jones et al., 2013; Plassmann et al., 2008; Field et al., 2014). However, the specific impacts of relatively low levels of nutrients from groundwater on aquatic habitats in dune systems have only recently been documented (Rhymes et al., 2014). As well as nutrients, diffuse inputs of dissolved organic matter (DOM) and micro-organisms into groundwater and surface waters also occur via runoff, field drainage and leaching, as a result of agricultural practices such as slurry and fertiliser application. Previously, these diffuse inputs have largely been characterised by using nutrients as a proxy (e.g. Vadas et al., 2007), although more recent studies are now examining diffuse sources and pathways by investigating pathogenic micro-organisms (Kay et al., 2008) and

704

J. Rhymes et al. / Science of the Total Environment 511 (2015) 703–710

characterising DOM by natural fluorescence (Hudson et al., 2007). To date there have been no studies combining chemical, fluorescent and microbial techniques to help decipher multiple diffuse sources and pathways. Excitation emission matrix fluorescence spectroscopy (EEMS) can be used to trace DOM from agricultural sources (Baker, 2002; Old et al., 2012). EEMS is sensitive enough to characterise fulviclike, humic-like and protein-like substances (Tryptophan-like and tyrosine-like) within the DOM to help characterise and quantify the extent of contamination by effluents from different sources (Hudson et al., 2007). Fulvic-like and humic-like substances are derived from the breakdown of plant material (Stedmon et al., 2003), whereas large inputs of tryptophan-like substances are associated with readily biodegradable material from sewage and farm waste slurry (e.g. Baker, 2001). Agricultural diffuse sources such as animal waste are characterised by high protein-like fluorescence with very high ratios of tryptophan-like to fulvic/humic-like fluorescence compared to stream waters (Baker, 2002); these ratios are sensitive enough to characterise inputs from different livestock animals such as pigs and sheep (Baker, 2002). Currently, the WHO Guidelines for Drinking Water Quality, adopted as standard in many countries, use total coliforms (TC), or specifically Escherichia coli (E. coli) a sub-group of faecal coliforms, as faecal indicators for the safety of water supplies. In some countries, such as The Netherlands and Denmark, groundwater is abstracted from sand dune systems to supply drinking water, indicating the importance of understanding the fate and occurrence of TC and E. coli within these systems (Smeets et al., 2009). The enumeration of TC and E. coli is also used as an indicator of water quality within the revised bathing water directive; E. coli counts greater than 10,000 counts per 100 mL and TC counts greater than 2000 counts per 100 mL would fail to meet the required standards in the directive (European Communitiy, 2006). Although the TC group includes the species E. coli, which is generally considered to be specific for faecal contamination, it also includes other genera such as Klebsiella and Citrobacter which are not necessarily of faecal origin and can emanate from alternative organic sources such as decaying plant materials and soils (WHO, 2006). While some E. coli represent enteric pathogens (Savageau, 1983), other strains of this species can grow and maintain populations in the environment if the conditions are suitable (Byappanahalli and Fujioka, 2004). Sources of E. coli include septic tanks, sewer lines, wastewater treatment plants, manure spreading on land, livestock and wildlife. These sources also contribute DOM and nutrient inputs. Despite advanced wastewater treatment efforts by water treatment companies, some UK bathing sites do not always produce full compliance with microbial standards (Crowther et al., 2002) due to other diffuse sources within catchments, resulting in a greater proportion of nonconformity due to agriculture. More than 150 different pathogens, associated with both environmental and human health risks can be found in livestock manure which can significantly increase bacterial loading to the subsurface, causing contamination within soils, groundwater and stream water (Gerba and Smith, 2005). The transport time and distance travelled by bacteria reaching the groundwater or streams depend on the rate at which bacteria are released from manure, the presence of preferential pathway networks within soil and the depth to the groundwater (Abuashour et al., 1994; Unc and Goss, 2003). The presence of TCs in surface or groundwater is usually considered evidence of recent faecal contamination, with E. coli remaining active for 16–45 days in the subsurface (Taylor et al., 2004). This study aims to use a combination of inorganic chemistry, DOM fluorescence and culturable E. coli and TC counts to evaluate the potential sources and pathways of nutrients and contamination to a sand dune site designated for its international nature conservation importance, known to be affected by nutrients from the surrounding agricultural land (Rhymes et al., 2014). Building on the previous study, a further year of bi-monthly sampling was carried out to separately assess the degree of potential contamination and

likely sources of contaminants from a) off-site sources entering the site from streams, b) off-site sources entering the site via runoff/ overland flow, c) groundwater flowing under the site, and d) onsite sources. 2. Materials and methods 2.1. Field monitoring strategy Aberffraw sand dune system is part of an internationally designated conservation site in the European Union Natura network, located on the southwest corner of the island of Anglesey in North Wales, UK (53°11′N, 4°27′W). It is designated for its dune habitats, in particular its dune slack wetlands and the rare plant and invertebrate species they support (Curreli et al., 2013). The site is in a low valley surrounded on three sides by agricultural land. The agricultural land is reseeded and fertilised pasture used for sheep and cattle grazing, with feed stations on land immediately adjacent to the south-east dune site edge (Fig. 1). Streams A and B (Fig. 1) drain this heavily fertilised agricultural and both lead onto the site. Flow in stream A is episodic and flows primarily in winter, compared with the permanent and faster flowing stream B. Annual long term average rainfall at the site is 847 mm (Stratford et al., 2013). There are a number of potential pathways by which nutrients and coliforms can enter the site; these include streams and ditches, surface runoff draining agricultural land and flowing onto the site, seepage of nutrients into the groundwater flowing under the site, and on-site sources such as grazing cattle and rabbits. Previous work has previously shown a nitrogen groundwater contamination gradient that extends into the site from the fertilised pastureland on the south east border with groundwater travelling in a south westerly direction (Fig. 1) (Rhymes et al., 2014). To determine the nature and pathways of the contamination, measurements were made bimonthly for a 12 month period (i.e. 6 sample periods) across streams, ditches, standing surface water and groundwater in dune slacks. Stream samples were collected from two streams (A and B) entering the site (Fig. 1). Samples were collected from upstream sampling points (A1 & B1) and downstream sampling points (A2 & B2) by dipping a clean collecting container into the surface flow. Groundwater samples within dune slacks were measured from fifteen groundwater monitoring piezometers across the site, installed to 2 m depth. Four piezometers (Fig. 1, triangles) aimed at looking at impacts from surface runoff. The remaining eleven aimed at evaluating potential gradients in the groundwater of water chemistry, natural fluorescence and TC and E. coli abundance with distance from the contamination sources on the south-east edge of the site (Fig. 1, squares). Samples were collected from the top 10 cm of the water table at each well using a sterilised pump and tubing, which was disinfected with Trigene and flushed three times with deionised water between samples into 250 mL sterile plastic bottles. During periods of inundation, for up to 4 months between November and February (Rhymes et al., 2014), when water tables were above ground level in certain slacks, samples of the standing water above the piezometer were taken. Groundwater depth was measured monthly at each piezometer. For fluorescence and E. coli measurements, sampling was conducted for 4 of the 6 sampling periods. 2.2. Water chemistry analysis Samples from piezometers and streams were analysed within 24 h of collection and stored in darkness at 5 °C prior to chemical analysis. In the laboratory groundwater pH was recorded for each sample which was then filtered through a 0.45 μm nylon syringe filter (Avonchem™). Dissolved inorganic anions (fluoride, chloride, nitrite, nitrate, phosphate and sulphate) and cations (sodium, ammonium, potassium, calcium and magnesium) were then quantified on an ion chromatograph (Metrohm, UK Ltd.). Dissolved inorganic nitrogen (DIN) was calculated as the sum of NO3-N, NO2-N and NH4-N. Total nitrogen (TN) and total carbon were analysed by thermal oxidation on a thermalox TOC/TN

J. Rhymes et al. / Science of the Total Environment 511 (2015) 703–710

705

Fig. 1. Map of Aberffraw dune system showing all piezometer and stream sampling points. Cross-hatched area represents designated site. The surrounding white area is agricultural land, predominantly pastureland. Redrawn from Ordnance Survey.

analyser. Total inorganic carbon (TIC) was measured within a TIC-reactor on a thermalox. Dissolved organic nitrogen (DON) was calculated by the difference between TN and calculated DIN and Dissolved Organic Carbon (DOC) was calculated by the difference between TC and TIC.

at 37 °C prior to colony counting: TCs generated purple colonies and E. coli produced blue colonies Harlequin™ E. coli/Coliform Medium.

2.5. Statistical analysis 2.3. Fluorescence analysis All samples were filtered in the field using 0.45 μm silver membrane filters (Steplitech) and stored in the dark at 5 °C prior to analysis. Analysis took place 48 h after collection at room temperature. Fluorescence measurements were obtained from a spectrophotometer (Variant Cary Eclipse) fitted with a xenon flash lamp using slit widths of 5 nm, an integration time of 12.5 ms and voltage of 700 V. Excitation wavelengths were varied from 200 to 400 nm in steps of 5 nm and emission wavelengths from 280 to 500 nm in steps of 2 nm. Post-processing was carried out using an R script described by Lapworth and Kinniburgh (2009) within the statistical package R. Absorbance was measured in a 1 cm cuvette on a UV–vis spectrophotometer (Varian Cary 50) at 1 nm intervals from 800 to 200 nm and SUVA254 was calculated by dividing absorbance at 254 nm by DOC concentration (Weishaar et al., 2003). Absorbance measurements were scatter corrected employing the method of Blough et al. (1993). All florescence data were corrected for instrument effects to account for lamp output, and corrected for inner filter effects using the corrected absorbance data (Lakowicz and Geddes, 1991). The data were reported in standard Raman units, which normalise the intensity by the area under the Raman peak between emission wavelengths of 380–410 for the excitation wavelength of 348 nm. 2.4. Enumeration of TC and E. coli Water samples were processed within 6 h of collection. E. coli and TC counts in water samples were determined in duplicate by filtering 20 mL of water through 0.2 μm cellulose nitrate filters (Whatman). Subsequently, cellulose nitrate filters were aseptically transferred onto Harlequin™ E. coli/Coliform Medium (Lab M). Culture plates were incubated for 22 h

All statistical analyses were performed using minitab v.16. Differences in stream water chemistry concentrations, fluorescence concentrations and E. coli and TC counts between upstream (A1, B1) and downstream (A2, B2) sampling points in two streams, A and B, were assessed using ANCOVA (Stevens, 1982), where the date of sample collection was used as a covariate to account for seasonal variation. In order to test for statistical differences in chemical, fluorescence and E. coli and TC count variables between streams, runoff and underlying groundwater gradients, annual means of all variables were analysed by grouping piezometers into three classes based on their distance from the south-east site edge (0–150 m, 150–300 m, 300– 450 m, excluding piezometers which were subject to run-off), grouping piezometers subject to runoff for March samples alone (see explanation below) and annual mean upstream sampling points from streams A and B (A1 & B1). Statistical tests used analysis of variance. Data that proved not normally distributed (Kolmogorov– Smirnov test) were transformed using a Johnson's transformation, which transforms the data to follow a normal distribution using the Johnson distribution system. In order to assess runoff inputs separately from any contributions from on-site sources, samples from piezometers subject to runoff (Fig. 1, triangles) were compared with all other piezometers (Fig. 1, squares) for samples collected in March, as this was the only month where all piezometers were subject to groundwater flooding. An ANCOVA (Stevens, 1982) was carried out on these two groups for all water chemistry, E. coli and TC counts and fluorescent spectroscopy variables, with distance from the south-east site edge as a covariate to account for potential underlying gradients in water chemistry due to other sources. Data that proved not normally distributed (Kolmogorov–Smirnov test) were transformed using a Johnson's transformation.

706

J. Rhymes et al. / Science of the Total Environment 511 (2015) 703–710

In order to assess underlying input gradients via the groundwater into the site, while accounting for runoff inputs, we separately analysed annual means of variables for all piezometers that were not subject to runoff (Fig. 1, squares). Relationships between annual means of all measurements with distance from the south-east site edge were investigated using linear regression. Data that proved not normally distributed (Kolmogorov–Smirnov test) were transformed using a Johnson's transformation. 3. Results When comparing the water chemistry, E. coli and TC counts and fluorescence among the main sources (Table 1), nitrate and DIN concentrations in runoff and streams were significantly higher than those found in groundwater samples from the 150–300 m and 300–450 m distance classes. There was no significant difference in nitrate concentrations between the three distance classes. However, significantly higher concentrations of DIN were observed in the 0–150 m class closest to the southeast site edge compared with the 300–450 m class. E. coli counts in the upstream sampling points of streams were significantly higher by an order of magnitude than in the runoff samples and in the groundwater at all distance classes (Table 1). TCs in streams were significantly higher than in the slacks subject to runoff, but were not significantly different from groundwater. There were no significant differences between sampling locations for DOC, phosphate, fulvic like, tryptophan like, TRP:FA and SUVA254. All fluorescent TRP:FA ratios measured within surface waters, groundwater and streams throughout the year did not exceed 1 R.U. and are described as uncontaminated drainage waters (Naden et al., 2010). 3.1. Stream nutrient and bacterial attenuation Upstream annual mean nitrate and DIN concentrations are significantly higher in stream A than in stream B, however E. coli counts are alike for both A and B streams for both upstream and downstream sampling points. Annual mean nitrogen concentrations upstream of streams entering the site are high at all stream sampling points (e.g. annual mean 12 mg/L of nitrate and 2.6 mg/L of DIN at B1) but are very high in A1, which drains from the south-east site edge into the site, where concentrations reached a maximum of 39 mg/L of nitrate in January. In stream A, the annual mean concentrations and counts for nitrate, DIN and E. coli were significantly higher upstream (A1) than those downstream (A2) (Fig. 2). By contrast, concentrations of nitrate and DIN in stream B (B1) did not decrease downstream to B2. No significance was found for fluorescence variables for all stream sampling points; TRP:FA mean concentrations were A1 = 0.217 ± 0.034 R.U., A2 = 0.280 ± 0.042 R.U., B1 = 0.457 ± 0.203 R.U. and B2 = 0.268 ±

0.042 R.U. Total nitrogen concentrations showed the same significant pattern as nitrate concentrations for all stream sampling points. 3.2. Runoff input In order to assess the contribution of off-site sources of contamination (cattle feed, overnight dunging, manure or slurry spreading) separately from any contribution by on-site sources (rabbits, dunging of cattle while grazing on-site), slacks only subject to groundwater flooding (Fig. 1, squares) were compared with those experiencing groundwater flooding in addition to runoff from neighbouring fields (Fig. 1, triangles), during a period when both sets of piezometers experienced surface inundation (Table 2). Significantly higher concentrations of nitrate and DIN were observed in those slacks exposed to runoff. However, there was no significant difference for DON, DOC, E. coli and TC counts, fluorescent variables and all other variables measured (Table 2). 3.3. Underlying gradients of nitrogen input via groundwater and on-site inputs In order to assess underlying input gradients via the groundwater into the site, while accounting for runoff inputs, we separately analysed annual means of variables for all piezometers that were not subject to runoff against distance from potential source area at the south-east site edge (Fig. 3). There were significant gradients of declining nitrate, DIN and DON concentrations into the site, away from their likely source at the south-east site edge (significant negative regression Fig. 3; a) Nitrate Coef = − 0.005, b) DON Coef = − 0.001 and c) DIN Coef = − 0.), with nitrate and DIN decreasing very strongly. A trend of linear decline was apparent for DOC concentrations but this was not significant (Fig. 3d). Fig. 4 shows the spatial pattern of E. coli and TC counts and tryptophan-like and fulvic-like fluorescence. Counts of E. coli (Figs. 3e and 4b) and TC (Fig. 4a) observed across the site showed no correlation with distance from the south-east site edge (Log E. coli counts/100 mL; max = 4.6, min = 0.00 and Log TC counts/100 mL; max = 4.85, min = 0.00). Tryptophan like fluorescence and fulvic like fluorescence (Fig. 3f and g) also showed no correlation with distance from the south-east south edge. However the TRP:FA ratio (Fig. 3h) showed a strong positive significant trend for increasing TRP:FA ratio away from the south-east site edge. 4. Discussion In this study we show that a combined analysis of nutrient concentrations, fluorescence and microbial abundance can help identify potential sources and pathways of nutrients and E. coli and TC inputs impacting a wetland site of international nature conservation

Table 1 Summary of annual mean water chemistry, E. coli and TC counts and fluorescence concentrations and counts for upstream sampling points for two streams (A and B), mean standing water for flooded slacks subject to runoff in March and for annual mean groundwater and standing water for distance classes (categorised piezometers located 0–150 m, 150–300 m and 300– 450 m away from the south east site edge, excluding piezometers subject to runoff). Values for each variable are expressed as mean ± standard error. Significant differences among classes are shown in bold; values with the same letter are not significantly different to each other. Variable

Stream

Groundwater and standing water Runoff

Water chemistry (mg/L)

Bacterial counts (Log 10 counts per 100 mL) Fluorescence (R.U.)

Absorbance (L mg−1 m−1)

Nitrate DIN DON DOC Phosphate E. coli TC Fulvic like Tryptophan like TRP:FA SUVA254

20.821 ± 9.763A 4.720 ± 2.209A 0.987 ± 0.459A 8.492 ± 13.903 0.750 ± 0.654 3.458 ± 0.033A 4.087 ± 0.197A 1.423 ± 0.092 1.141 ± 0.030 0.810 ± 0.036 0.043 ± 0.011

7.519 ± 1.556A 1.702 ± 0.352A 0.658 ± 0.042AB 24.600 ± 13.903 0.009 ± 0.003 0.000 ± 0.000B 2.559 ± 0.169B 1.491 ± 0.003 1.167 ± 0.002 0.782 ± 0.001 0.061 ± 0.046

Distance from site edge (m) 0–150

150–300

300–450

3.197 ± 1.516AB 0.756 ± 0.346AC 0.430 ± 0.076AB 14.250 ± 0.966 0.020 ± 0.003 0.151 ± 0.151B 2.863 ± 0.083AB 1.321 ± 0.048 1.114 ± 0.015 0.844 ± 0.019 0.018 ± 0.003

0.033 ± 0.019B 0.050 ± 0.019BC 0.322 ± 0.088B 7.789 ± 4.878 0.012 ± 0.001 0.403 ± 0.242B 2.793 ± 0.355AB 1.367 ± 0.078 1.123 ± 0.024 0.838 ± 0.027 0.020 ± 0.003

0.018 ± 0.007B 0.017 ± 0.003B 0.249 ± 0.034B 8.602 ± 1.863 0.015 ± 0.005 0.285 ± 0.133B 2.641 ± 0.431AB 1.251 ± 0.041 1.097 ± 0.016 0.888 ± 0.017 0.015 ± 0.005

J. Rhymes et al. / Science of the Total Environment 511 (2015) 703–710

b

a

c

bcc

b

c

10 0 a

8 6 4

c

2 0

bc

b

A1 A2 B1 B2 2

A1 1 A2 A B1 B2

Stream samp s plingg poiint

Stre eam sam mpling po oint

Mean LLog E.coli per 100 ml) M E li (Counts (C

40 35 5 30 0 25 5 20 0 15 5 10 0 5 0

Mean DIN (mg/L)

Mean nitrate ((mg/L) g/ )

a

4 3.5 3 2.5 2 1.5 1 0.5 0

707

a

a

aab b

A1 A2 A B1 1 B2 Stream m sam mpliing p pointt

Fig. 2. Annual mean concentrations for nutrients and E. coli counts of two streams (A and B) entering the site at upstream and downstream sampling points, showing: a) Nitrate b) DIN and c) Log E. coli counts. Letters denote significant differences between stream sampling points A1 & B1 upstream and A2 & B2 downstream — see Fig. 1.

importance. The multiple pathways and the fate of nutrients and E. coli and TC to the site are summarised in Fig. S1 in the Supplementary Material.

4.1. Off-site sources (fertiliser, cattle dung, slurry and fertilisers applied to fields) 4.1.1. Streams Streams A and B have similar annual mean E. coli abundances as they enter the site, suggesting that both streams have similar faecal inputs from grazers on adjacent pastureland. By contrast, nitrate concentrations in Stream A, which drains the pastureland and flows onto the sand dune site, are significantly higher than those in stream B and previous studies have shown that they can exceed the 50 mg/L threshold for designation of a nitrate vulnerable zone by the UK Environment Agency (Rhymes et al., 2014; Environment Agency, 2012). This implies that stream A has additional N inputs compared to stream B; these are likely to be from fertilisers leaching from the steep sloped pastureland adjoining the stream. This is a result of the lack of significant difference in TRP:FA ratio between the two streams suggesting a common contamination source from animal dung, with differences in nutrient loading due to fertiliser inputs only. The attenuation of nitrate concentrations in stream A is probably caused by a combination of processes, including in-stream microbial denitrification, plant uptake, seepage through the river bank, dilution during high groundwater levels and transient storage (Mulholland et al., 2008). The lack of nutrient attenuation in stream B may be due to the

Table 2 Summary of selected chemical, fluorescent and E. coli and TC count variables for piezometers subject to both runoff and groundwater flooding and wells subject to groundwater flooding alone in March. Values for each variable are expressed as mean ± standard error. Significant differences between groups of slacks are shown in bold and denoted by letters. Variable

Water chemistry (mg/L)

Bacterial counts (Log 10 counts per 100 mL) Fluorescence spectroscopy (R.U.) Absorbance (L mg−1 m−1)

Slacks subject to:

Nitrate DIN DON DOC Phosphate E. coli TC Fulvic like Tryptophan like TRP:FA Suva254

Runoff and groundwater flooding

Groundwater flooding

7.519 ± 1.556A 1.702 ± 0.352A 0.658 ± 0.042 11.100 ± 4.700 0.014 ± 0.009 0.000 ± 0.000 3.036 ± 0.301 1.491 ± 0.003 1.167 ± 0.002 0.782 ± 0.001 0.061 ± 0.046

0.676 ± 0.386B 0.181 ± 0.0920B 0.505 ± 0.072 19.117 ± 2.023 0.024 ± 0.008 0.841 ± 0.351 3.370 ± 0.086 1.515 ± 0.074 1.172 ± 0.023 0.785 ± 0.024 0.018 ± 0.003

absence of nitrophilous species such as Phragmites australis (Stamati et al., 2010) and the faster stream flow in stream B which reduces the water-sediment contact (Peterson et al., 2001) and reduces the ability for microbes to assimilate nitrogen from the water column (Grimm and Fisher, 1989). Similarly E .coli counts decrease from upstream to downstream sampling points in stream A but not stream B. Studies have shown that within sediment there are higher populations of TCs than the overlaying water (Smith et al., 2008) as sediments serve as a hospitable environment for bacterial survival due to the availability of organic matter (Jamieson et al., 2005) suggesting that E. coli may be being deposited into the stream and incorporated within the sediment. Subsequently during storm events bacteria can be re-suspended into the water column and continue to flow downstream (Jeng et al., 2005), posing a potential threat to bathing waters on the sandy beach at Aberffraw at the mouth of the river Ffraw since mean E. coli colonies within both streams exceed 2000 colonies per 100 mL which would fail to meet the required standards for the EU bathing water directive (2006).

4.1.2. Surface runoff The nature of the nitrate contamination contributing via runoff is also likely to be due to applied fertilisers on the south-east pastureland, rather than nutrients from slurry and dung. Nitrate concentrations are eleven times higher in slacks subject to runoff and groundwater flooding compared to piezometers subject to groundwater flooding alone, whereas the TRP:FA fluorescence ratios were b 1 R.U. which are described as uncontaminated drainage waters compared to the described slurry TRP:FA fluorescence ranging from 2–5 R.U. (Naden et al., 2010). In support of this, negligible E. coli and TC counts were observed in slacks subject to runoff which would have been expected if the contamination resulted from slurry application (Thurston-Enriquez et al., 2005). Surface runoff events are sporadic as they are caused by heavy periods of rainfall, nevertheless while nutrient concentrations are high during runoff events and in the subsequent standing surface water, the concentrations in groundwater return to low concentrations once the runoff ceases. This may be a result of denitrification caused by the increased availability of nitrate and anaerobic conditions (Mulvaney et al., 1997), or due to plant and microbial uptake highlighting the function of dune systems in filtering nutrients. However, the initial nitrate and DIN concentrations in the standing water are much higher than the levels of 0.2 mg DIN L−1 above which biological effects have been determined previously at the site (Rhymes et al., 2014), suggesting adverse impacts on the site due to nutrients from this source. In addition, the total flux of nitrogen entering the site as a result of these runoff events is also unknown. Calculating this input may help harmonise dose-related critical load approaches to damage (Bobbink and Hettelingh, 2010) with concentration based methods used in aquatic systems (Camargo and Alonso, 2006).

708

J. Rhymes et al. / Science of the Total Environment 511 (2015) 703–710

Fig. 3. Relationships between annual mean groundwater a) nitrate, b) DON, c) DIN, d) DOC, e) E. coli, f) tryptophan-like, g) fulvic-like, and h) TRP:FA with distance from south-east site edge. Trendline for all variables is linear.

4.1.3. Groundwater Nutrients, likely to be from ammonium nitrate fertiliser, are entering the site by leaching through sandy agricultural soils with a low water holding capacity (Skiba and Wainwright, 1984), and subsequently flowing under the site via the groundwater. As a result a nitrate, DIN and DON groundwater gradient was found independent of runoff influence, confirming that the gradients of elevated nutrients in groundwater and soils observed in an earlier study were not due to surface flooding (Rhymes et al., 2014). There was no gradient in E. coli and TC counts within the groundwater under the site, suggesting that the higher concentrations of E. coli and TCs originating in the south-east pasture and observed in the streams are probably filtered out by the

sandy soils during recharge transit in the subsurface before reaching the site (Price et al., 2013). 4.2. On-site sources (rabbits, cattle dung) E. coli and TCs were found across the site at low levels, but showed no relationship with the distance from the south-east site edge. Similarly, the tryptophan like fluorescence showed no relationship with the distance from the south-east site edge. This suggests that the main source of groundwater E. coli and TCs derive from on-site cattle and rabbits. Therefore, in addition to streambed seepage, runoff and underlying groundwater nutrient inputs, there are nutrient and E. coli and TC inputs

J. Rhymes et al. / Science of the Total Environment 511 (2015) 703–710

709

Fig. 4. Spatial variation of a) Log E. coli counts/100 mL, b) Log total coliform counts/100 mL, c) tryptophan-like fluorescence (R.U.) and d) fulvic-like fluorescence (R.U.) for streams (annual mean), piezometers subject to runoff (March counts and concentrations) and piezometers (annual mean).

from on-site grazers such as cattle and rabbits. Despite these small-scale on-site inputs, all piezometers across the site meet the mandatory bathing water directive standards of 2000 E. coli counts per 100 mL.

or other animal sources of contamination, and methods that do so are expensive (Field and Samadpour, 2007). Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.scitotenv.2014.12.085.

5. Conclusions The combination of chemical, fluorescent and microbial techniques has helped identify potential nutrient sources from fertilisers and grazers (Baker, 2002). The findings of this study suggest that nutrients are being attenuated and processed within the site thereby providing a valuable ecosystem service. However at the same time, the influx of nutrients is likely to have an adverse effect on the dune slack ecology, with impacts on plant community composition (Rhymes et al., 2014). The analysis of surface waters, slacks and piezometers across the site has allowed the differentiation between the input pathways of streams, runoff events, underlying groundwater nutrient gradients and on-site grazing inputs. While the study was able to distinguish multiple pathways, the full potential of the techniques to differentiate between livestock sources (e.g. sheep, cattle, pigs) was not explored in this study. This combination of techniques provides an approach which could allow for a detailed understanding of nutrient contamination sources and pathways relatively cheaply. Such information is key for designing successful management plans to reduce the inputs of contaminants which might be having detrimental effects on sites of conservational value. It could also be implemented for other applications such as tracking faecal sources within bathing waters and fisheries zones, as currently the standard enumeration of FIB does not distinguish between human

References Abuashour, J., Joy, D.M., Lee, H., Whiteley, H.R., Zelin, S., 1994. Transport of microorganisms through soil. Water Air Soil Pollut. 75, 141–158. Baker, A., 2001. Fluorescence excitation-emission matrix characterization of some sewage-impacted rivers. Environ. Sci. Technol. 35, 948–953. Baker, A., 2002. Fluorescence properties of some farm wastes: implications for water quality monitoring. Water Res. 36, 189–195. Blough, N.V., Zafiriou, O.C., Bonilla, J., 1993. Optical-absorption spectra of waters from the Orinoco River outflow — terrestrial input of colored organic-matter to the Caribbean. J. Geophys. Res. Oceans 98, 2271–2278. Bobbink, R., Hettelingh, J.-P., 2010. Review and revision of empirical critical loads and dose–response relationships. Proceedings of an Expert Workshop, Noordwijkerhout, pp. 23–25. Byappanahalli, M., Fujioka, R., 2004. Indigenous soil bacteria and low moisture may limit but allow faecal bacteria to multiply and become a minor population in tropical soils. Water Sci. Technol. 50, 27–32. Camargo, J.A., Alonso, A., 2006. Ecological and toxicological effects of inorganic nitrogen pollution in aquatic ecosystems: a global assessment. Environ. Int. 32, 831–849. Crowther, J., Kay, D., Wyer, M.D., 2002. Faecal-indicator concentrations in waters draining lowland pastoral catchments in the UK: relationships with land use and farming practices. Water Res. 36, 1725–1734. Curreli, A., Wallace, H., Freeman, C., Hollingham, M., Stratford, C., Johnson, H., Jones, L., 2013. Eco-hydrological requirements of dune slack vegetation and the implications of climate change. Sci. Total Environ. 443, 910–919. Environment Agency, 2012. Method statement for nitrate vulnerable zone review. Surface Waters Environment Agency Report to Defra and Welsh Government — Supporting Paper for the Implementation of the Nitrates Directive 2013–2016.

710

J. Rhymes et al. / Science of the Total Environment 511 (2015) 703–710

European Community, 2006. Directive 2006/7/EC of the European Parliament and of the Council of 15 February (2006) concerning the management of bathing water quality and repealing Directive 76/160/EEC. Official Journal of the European Union, L64/ 37–51. Field, K.G., Samadpour, M., 2007. Fecal source tracking, the indicator paradigm, and managing water quality. Water Res. 41, 3517–3538. Field, C.D., Dise, N.B., R.J., P., Britton, A.J., Emmett, B.A., Helliwell, R.C., Hughes, S., Jones, L., Lees, S., Leake, J.R., Leith, I.D., Phoenix, G.K., Power, S.A., Sheppard, L.J., Southon, G.E., Stevens, C.J., Caporn, S.J.M., 2014. The role of nitrogen deposition in widespread plant community change across semi-natural habitats. Ecosystems 17, 864–877. Fowler, D., O'Donoghue, M., Muller, J., Smith, R., Dragosits, U., Skiba, U., Sutton, M., Brimblecombe, P., 2005. A chronology of nitrogen deposition in the UK between 1900 and 2000. Water Air Soil Pollut. Focus 4, 9–23. Galloway, J.N., Cowling, E.B., 2002. Reactive nitrogen and the world: 200 years of change. Ambio 31, 64–71. Gerba, C.P., Smith, J.E., 2005. Sources of pathogenic microorganisms and their fate during land application of wastes. J. Environ. Qual. 34, 42–48. Grimm, N.B., Fisher, S.G., 1989. Stability of periphyton and macroinvertebrates to disturbance by flash floods in a desert stream. J. N. Am. Benthol. Soc. 8, 293–307. Hooda, P.S., Edwards, A.C., Anderson, H.A., Miller, A., 2000. A review of water quality concerns in livestock farming areas. Sci. Total Environ. 250, 143–167. Hudson, N., Baker, A., Reynolds, D., 2007. Fluorescence analysis of dissolved organic matter in natural, waste and polluted waters — a review. River Res. Appl. 23, 631–649. Jamieson, R., Joy, D.M., Lee, H., Kostaschuk, R., Gordon, R., 2005. Transport and deposition of sediment-associated Escherichia coli in natural streams. Water Res. 39, 2665–2675. Jeng, H.W.C., England, A.J., Bradford, H.B., 2005. Indicator organisms associated with stormwater suspended particles and estuarine sediment. J. Environ. Sci. Health. Part A Toxic/Hazard. Subst. Environ. Eng. 40, 779–791. Jones, L., Nizam, M., Reynolds, B., Bareham, S., Oxley, E., 2013. Upwind impacts of ammonia from an intensive poultry unit. Environ. Pollut. 180, 221–228. Kay, D., Crowther, J., Fewtrell, L., Francis, C.A., Hopkins, M., Kay, C., McDonald, A.T., Stapleton, C.M., Watkins, J., Wilkinson, J., Wyer, M.D., 2008. Quantification and control of microbial pollution from agriculture: a new policy challenge? Environ. Sci. Pol. 11, 171–184. Lakowicz, J.R., Geddes, C.D., 1991. Topics in Fluorescence Spectroscopy. Springer. Lapworth, D.J., Kinniburgh, D.G., 2009. An R script for visualising and analysing fluorescence excitation–emission matrices (EEMs). Comput. Geosci. 35, 2160–2163. Mulholland, P.J., Helton, A.M., Poole, G.C., Hall, R.O., Hamilton, S.K., Peterson, B.J., Tank, J.L., Ashkenas, L.R., Cooper, L.W., Dahm, C.N., Dodds, W.K., Findlay, S.E.G., Gregory, S.V., Grimm, N.B., Johnson, S.L., McDowell, W.H., Meyer, J.L., Valett, H.M., Webster, J.R., Arango, C.P., Beaulieu, J.J., Bernot, M.J., Burgin, A.J., Crenshaw, C.L., Johnson, L.T., Niederlehner, B.R., O'Brien, J.M., Potter, J.D., Sheibley, R.W., Sobota, D.J., Thomas, S.M., 2008. Stream denitrification across biomes and its response to anthropogenic nitrate loading. Nature 452, 202-U46. Mulvaney, R.L., Khan, S.A., Mulvaney, C.S., 1997. Nitrogen fertilizers promote denitrification. Biol. Fertil. Soils 24, 211–220. Naden, P.S., Old, G.H., Eliot-Laize, C., Granger, S.J., Hawkins, J.M.B., Bol, R., Haygarth, P., 2010. Assessment of natural fluorescence as a tracer of diffuse agricultural pollution from slurry spreading on intensely-farmed grasslands. Water Res. 44, 1701–1712. Novotny, V., 1999. Diffuse pollution from agriculture — a worldwide outlook. Water Sci. Technol. 39, 1–13. Oakes, D.B., Young, C.P., Foster, S.S.D., 1981. The effects of farming practices on groundwater quality in the United-Kingdom. Sci. Total Environ. 21, 17–30.

Old, G.H., Naden, P.S., Granger, S.J., Bilotta, G.S., Brazier, R.E., Macleod, C.J.A., Krueger, T., Bol, R., Hawkins, J.M.B., Haygarth, P., Freer, J., 2012. A novel application of natural fluorescence to understand the sources and transport pathways of pollutants from livestock farming in small headwater catchments. Sci. Total Environ. 417, 169–182. Peterson, B.J., Wollheim, W.M., Mulholland, P.J., Webster, J.R., Meyer, J.L., Tank, J.L., Marti, E., Bowden, W.B., Valett, H.M., Hershey, A.E., McDowell, W.H., Dodds, W.K., Hamilton, S.K., Gregory, S., Morrall, D.D., 2001. Control of nitrogen export from watersheds by headwater streams. Science 292, 86–90. Plassmann, K., Brown, N., Jones, M.L.M., Edwards‐Jones, G., 2008. Can atmospheric input of nitrogen affect seed bank dynamics in habitats of conservation interest? The case of dune slacks. Appl. Veg. Sci. 11, 413–420. Price, W.D., Burchell, M.R., Hunt, W.F., Chescheir, G.M., 2013. Long-term study of dune infiltration systems to treat coastal stormwater runoff for fecal bacteria. Ecol. Eng. 52, 1–11. Rhymes, J., Wallace, H., Fenner, N., Jones, L., 2014. Evidence for sensitivity of dune wetlands to groundwater nutrients. Sci. Total Environ. 490C, 106–113. Savageau, M.A., 1983. Escherichia coli habitats, cell-types, and molecular mechanisms of gene-control. Am. Nat. 122, 732–744. Skiba, U., Wainwright, M., 1984. Nitrogen transformation in coastal sands and dune soils. J. Arid Environ. 7, 1–8. Smeets, P., Medema, G., Van Dijk, J., 2009. The Dutch secret: how to provide safe drinking water without chlorine in the Netherlands. Drinking Water Eng. Sci. 2, 1–14. Smith, J., Edwards, J., Hilger, H., Steck, T.R., 2008. Sediment can be a reservoir for coliform bacteria released into streams. J. Gen. Appl. Microbiol. 54, 173–179. Stamati, F.E., Chalkias, N., Moraetis, D., Nikolaidis, N.P., 2010. Natural attenuation of nutrients in a Mediterranean drainage canal. J. Environ. Monit. 12, 164–171. Stedmon, C.A., Markager, S., Bro, R., 2003. Tracing dissolved organic matter in aquatic environments using a new approach to fluorescence spectroscopy. Mar. Chem. 82, 239–254. Stevens, J.P., 1982. The analysis of covariance and alternatives — Huitema, Be. Contemp. Psychol. 27, 473-473. Stratford, C.J., Robins, N.S., Clarke, D., Jones, L., Weaver, G., 2013. An ecohydrological review of dune slacks on the west coast of England and Wales. Ecohydrology 6, 162–171. Taylor, R., Cronin, A., Pedley, S., Barker, J., Atkinson, T., 2004. The implications of groundwater velocity variations on microbial transport and wellhead protection — review of field evidence. FEMS Microbiol. Ecol. 49, 17–26. Thurston-Enriquez, J.A., Gilley, J.E., Eghball, B., 2005. Microbial quality of runoff following land application of cattle manure and swine slurry. J. Water Health 3, 157–171. Unc, A., Goss, M.J., 2003. Movement of faecal bacteria through the vadose zone. Water Air Soil Pollut. 149, 327–337. Vadas, P.A., Harmel, R.D., Kleinman, P.J.A., 2007. Transformations of soil and manure phosphorus after surface application of manure to field plots. Nutr. Cycl. Agroecosyst. 77, 83–99. Weishaar, J.L., Aiken, G.R., Bergamaschi, B.A., Fram, M.S., Fujii, R., Mopper, K., 2003. Evaluation of specific ultraviolet absorbance as an indicator of the chemical composition and reactivity of dissolved organic carbon. Environ. Sci. Technol. 37, 4702–4708. WHO, 2006. Guidelines for Drinking-water Quality: First Addendum to Volume 1, Recommendations. World Health Organization. Withers, P.J.A., Jarvie, H.P., Hodgkinson, R.A., Palmer-Felgate, E.J., Bates, A., Neal, M., Howells, R., Withers, C.M., Wickham, H.D., 2009. Characterization of phosphorus sources in rural watersheds. J. Environ. Qual. 38, 1998–2011.